MARCH 1961
SOME NEW FUMUIC ACID DERIVATIVES
*
697
[CONTRIBUTIONFROM THE RESEARCH LABORATORIES OF J. T. BAKERCHEMICAL COMPANY]
Some New Fumaric Acid Derivatives. Preparation of Mixed Fumarates and Thiolfumarates PAUL G. CAMPBELL, GENE SUMRELL,
AND
CHARLES H. SCHRAMM
Received May 16, 1980 Several new bis-aryl fumarata were prepared by the Schotten-Baumann method. Attempta to modify the chemical and physical properties of these compounds led to the synthesis of certain new alkyl fumaryl Chlorides by a convenient method. This involved the reaction of maleic anhydride with an alcohol, followed by treatment with thionyl chloride. The acid chlorides were used aa intermediates in the preparation of a variety of mixed fumarate and thiolfumarate eatera, and of some new fumaramates.
In the framework of a larger program, we prepared several halogenated bisaryl fumarate esters to investigate their utility in conveying flame retardant properties to copolymers. These compounds were prepared by a Schotten-Baumann reaction utilizing a halogenated phenol and fumaryl chloride.'I2 The high melting points and slight solubility of these materials in organic liquids led us to attempt to obtain materials with more desirable physical properties by incorporating an alkyl group on one end of the fumarate structure. Such mixed esters do not appear to have been reported in the literature. We are reporting here the preparation of these and related compounds. A promising approach to the synthesis of such compounds appeared to be the reaction of a phenol with an alkyl half-ester, half-acid chloride of the fumaric acid. A convenient method of preparing ethyl fumaryl chloride in good yield has been reported by Eisner, Elvidge, and Linstead, a involving the isomerization of ethyl hydrogen maleate to the fumarate, followed by reaction with thionyl chloride. These workers also prepared methyl fumaryl chloride similarly, though in lower yield. We attempted to apply the method to the preparation of n-propyl fumaryl chloride and obtained poor results. This led to an attempt to convert crude ncl -
II
0
1
CH-C-OR SOCl 1 c
propyl hydrogen maleate (from the alcohol and maleic anhydride) to n-propyl fumaryl chloride in one step by refluxing with a small excess of thionyl chloride. The method was very successful, giving an 80% yield of the desired acid chloride which was easily purified by fractional distillation. Table I gives the data on the alkyl fumaryl chlorides prepared in this manner.' The lower yield of the ipropyl derivative may, in part, have been due to the incomplete esterification of i-propyl alcohol by maleic anhydride. This difference in reactivity of aaleic anhydride with primary and secondary alcohols has previously been observed by Siegel and Moran.46 In the course of our work we used these alkyl fumaryl chlorides to prepare a number of alkyl aryl fumarates, the aryl group being a halogenated phenyl group. As expected, the compounds were much lower melting than the bisaryl fumarates, several of them being liquids a t room temperature. We also allowed these acid chlorides to react with several primary and secondary amines yielding fumaramates. These were solids with suitable solubility properties for polymerization studies with liquid comonomers. The acid chlorides were also allowed to react with various mercaptans, yielding mixed thiolfumarates. As expected, these compounds were also considerably lower melting and more soluble than the dithiolfumarates previously reported from these
(4)Attempta to extend this method to the preparation of either methyl or ethyl fumaryl chloride gave mixtures CH-C CH-$!-OH 0 which were largely the desired acid chlorides. However, fractional distillation waa not succeasful in removing traces of 0 the dialkyl fumarates and possibly other impurities. Yields of 40 to 50% of product of >90% purity were obtained. (5) E. F. Siegel and M. K. Moran, J . Am. Chem. Soc., 69, 1457 (1947). 0 (6)When this reaction waa carried out between tbutyl (1) For the similar reaction of methacrylyl chloride with alcohol and maleic anhydride, maleic acid was obtained in thiolphenols, see G. Sumrell, G. E. Ham, and E. D. Horn- 97.8% yield. The maleic anhydride dissolved upon heating baker, J. Am. Chem. SOC.,80,2509 (1958). on the steam bath with tbutyl alcohol, but after three (2)For the simiiar preparation of aryl methacrylates, quarters of an hour a vigorous exothermic reaction occurred see (a) 5. Patai, M. Bentov, and M. E. Reichmann, J. Am. with gaa evolution, and the residual maleic acid solidified. Chem. Soc., 74,845 (1952);and (b) G.Sumrell, P. G. C a m p The maleic acid waa identified by elemental analysia and bell, C. H. Schramm, and G. E. Ham, J. Am. Chem. SOC.,81, melting point determinations. The evolved gaa decolorized 4310 (1959). bromine in carbon tetrachloride waa presumably isobutylene. (3) U.Eisner,.J. A. Elvidge, and R. P. Linstead, J. C M . It is possible that tbia reaction may be of value as a synthetic Soc., 1501 (1951). tool for converting tertiary alcohols to olefins. loo", 4 hr;
698
CAMPBELL, SWMRELL, AND SCHRAMM
VOL.
26
TABLE I 0
R0-d lE' ALKYLFUMABYL CHLORIDES Yield,
R
B.P.
ny
%
wC~HVGC8H.r WC~HO*CsHii-
99-100/15 mm. 8&90/12 mm. 102-104/9 mm. 120.b122/13 mm.
1.4599 1.4540 1.4610 '1.4612
80.2 46.8 82.1 83.4
Formula CvH&10, (3,HpClOa CsHiiClOj CeHisClOs
Iaborat~ries.~fi Table I1 lists the properties of the fumaric acid derivatives prepared.
C 47.60 47.60 50.40 52.82
Calcd. H Halogen 5.14 5.14 5.82 6.40
20.08 20.08 18.60 17.33
Found C
H
Halogen
47.73 47.68 50.55 52.88
5.26 5.30 5.96 6.51
20.07 20.23 18.60 17.17
Anal. Calcd. for C16H8C1,04:C, 47.32:; H, 1.98; C1, 34.93. Found: C, 47.28; H, 1.97; C1, 34.71. Bie( 9,,$,6-triehlorophenyl)fumurate waa similarly prepared and recrystallized; yield, 96.3 g. (40%), m.p. 205-206'. Anal. Calcd. for Ci6&cl60~: C, 40.46; HI 1.27; CI, 44.79. EXPERIMENTAL Found: C, 40.53; H, 1.32; C1,44.62. n-Propyl f u m r y l cMode.1Q A sample of crude n-propyl Boiling points and melting points are uncorrected. Unless otherwise indicated, distillations were carried out through an hydrogen maleate waa prepared by warming a mixture of 1 mole each of maleic anhydride and n-propyl alcohol on the 80-cm. Podbielniak-type column. Starting materials. The phenols, amines, mercaptans, and steam bath until a syrupy liquid resulted. To this mixture maleic anhydride were commercial materials and were used was added dropwise and with stirring 1.25 moles of thionyl without purification. The thionyl chloride was distilled from chloride. The temperature was maintained below 45' by quinoline before use. Methyl and ethyl fumaryl chloride cooling. After the addition waa complete, the mixture waa were prepared essentially as described by Eisner, Elvidge gradually heated to 100' in ca. 2 hr., and maintained a t this temperature for an additional 4 hr. The dark mixture was and Linstead.8 Bis(4-bromophenyl) fumarate. The preparation of bik(4- fractionally distilled, yielding 141.5 g. (80%) of product a t bromophenyl) fumarate illustrates the method used for the 99-100" (15 mm.), n y 1.4599. The other alkyl fumaryl chlorides (Table I) were prepared remaining bisaryl fumarate esters. One mole of Pbromophenol was dissolved in 1000 ml. of 5% sodium hydroxide in by the same procedure. Methyl 9,4,6-trichlorophenyl fumarate. A mixture of 43.4 a 2 1. three necked flask fitted with a dropping funnel, stirrer and thermometer. The solution was maintained at 0' with g. (0.22 mole) of 2,4,5-trichlorophenol, 10 g. (0.25 mole) of an acetone-ice bath while 0.5 mole of fumaryl chloride was sodium hydroxide and 200 ml. of water was cooled to 0", added dropwise. Stirring was continued for 1 hr. Then the and 29.7 g. (0.2 mole) of methyl fumaryl chloride was added solid was filtered, washed well with water, and air dried. dropwise with stirring. The mixture was allowed to stir Recrystallization from a mixture of benzene and acetone gave -era1 hours. The reaction mixture consisted of an upper 96.3 g. (44%) of product of m.p. 174-175 (lit.,9 m.p. 174"). aqueous phase and a lower solid phase. The solid was filBis($4-dibrmnophen& fumarate was similarly prepared tered, washed with water, and recrystallized from isopropyl alcohol; yield, 18.0 g. (29.9%), m.p. 36-37". This general and recrystallized from benzene-acetone:, vield, - procedure " . 110.7 E. waa used for the preparation of the remaining alkyl (38%), m.p. 215-216'. Anal. Calcd. for ClsH8Br404:C, 32.91; H, 1.38; Br, 54.75. aryl fumarates (Table 11). Ethgl N-morphalinojumarate. The preparation of this comFound: C, 32.88: H. 1.50: Br. 54.65. Bis(8,4&chl&ophenyl) fumarate was similarly prepared pound illustrates the method used to synthesize the fumaramates described in Table 11. Morpholine (0.23 mole), disand recrystallized; yield 87.7 g. (43%), m.p. 186-187. solved in 200 ml. of ethylene dichloride, was placed in a 500 (7) G. Sumrell, M. Zief, E. J. Huber, C. H. Schramm, ml. three necked flask fitted with a stirrer, dropping funnel, and G. E. Ham, J. Am. Chem. SOC., 81, 4313 (1959). and thermometer. The mixture was cooled to 0" and ethyl (8)It was pointed out in Reference 7 that the dithiol fumaryl chloride (0.1 mole) was added dropwise. The mixture fumarates with aryl groups in the ester functions were yellow was allowed to stir overnight. The organic solution was in color. Even di-tbutyl dithiolfumarate and some simpler washed with dilute phosphoric acid and then with brine' thiol esters such aa pentachlorochlorophenyl thiolacetate and until the washings were neutral, The ethylene dichloride W&O pentachlorophenyl thiolmethacrylate proved to be yellow removed by flash evaporation, The semisolid residue was solids. Similarly, the thiolfumarates described in this paper recrystallized from ligroin-isopropyl alcohol, giving 16.6 g. were yellow in color when an aryl group was in the thiol (78%) of product, m.p. 72.5-73'. ester function. However, 0-methyl S-t-butyl thiolfumarate, 0-Methyl S-methyl thiolfumarate. Methyl mercaptan was though yellow as a liquid, on cooling formed a colorless bubbled into 100 ml. of ice cold anhydrous l,>dimethoxycrystalline solid which melted at 24'. 0-Ethyl S-&butyl ethane until the weight increased 24 g. (0.5 mole). This soluthiolfumarate was a yellow oil which froze to a colorless solid (10) n-Propyl hydrogen fumarate was prepared in So% in Dry Ice. The reason for the yellow color in these compounds, seem- yield on a 1 mole scale, following the procedure given in refingly lacking in chromophoric groups, is speculative. Dr. erence 3 for ethyl hydrogen fumarate. This compound was E. M. Kosower, University of Wisconsin, has suggested in a converted to the acid chloride in only 20% yield using phosprivate communication that the color may be due to inter- phorus trichloride. The use of thionyl chloride did not inor intramolecular charge-transfer transitions. We are a t prove the yield. For n-propyl hydrogen fumarate, m.p. 4849'. Anal. Calcd. for C&QO~:C, 53.16; HI 6.37. Found: present not able to offer a more suitable explanation. C, 53.16; H, 6.43. (9) R. Anschutz, Chem. Ber., bOB, 1320 (1927). '
MARCH
1961
699
SOME NEW FUMARIC ACID DERIVATIVES
mu)
ml-
093. M d A d
1 - 4 0
-4m
3 M W 3-N
32
??N.
25
??
I
0-13
\u/ I1
x
/O
\
y=c I
2
g$
-44
h
4
s
I
L I
dl?
2s
700
HASEK, ELAM, MARTIN, AND NATIONS
tion was mixed with 74.4 g. (0.5mole) of methyl fumaryl chloride in a 500 ml. three necked h k equipped with stirrer, dropping funnel, and thermometer. The temperature was maintained at 0-5" by cooling with an acetone ice bath while 40 ml. of pyridine was added dropwise with stirring. After 0.5 hr. of additional stirring, 200 ml. of water was added. The mixture waa cooled back to 0' and filtered. The brown solid was washed thoroughly with water and air dried; crude yield, 57 g. (71%), m.p. 50-53". Recrystalliiation from a mixture of 120 ml. of isopropyl alcohol and 120 ml. of water gave 47 g. of long colorleas blades, m.p. 63-54", unchanged by further recrystallization. 0-Methyl S-isopropyl thiolfumarak waa prepared in a similar manner. However, the crude product separated as an oily lower phase. This was extracted into 50 ml. of chloroform, washed with water, and dried over sodium sulfate. The material was then purified by fractional distillation. 0-Methyl S-tht?/l thiolfumarate. The synthesis of this compound describes the method used for the preparation of
[CONTRIBUTION FROM
THE
VOL.
26
the remaining thiolfumarates. A solution of 36 g. (0.4mole) of tbutyl mercaptan in 400 ml. of 5% aqueous sodium hydroxide was maintained at 10-15' by cooling while 59.5 g. (0.4mole) of methyl fumaryl chloride was added dropwise with stirring. An oil separated to the bottom when stirring waa stopped. The mixture was stirred for 3 hr. with cooling to -5" in an unsucceeaful attempt to induce crystallization. Then the oil was extracted into 50 ml. of chloroform, washed with water, and dried over sodium sulfate. Fractional distillation gave 32 g. (40%) of light yellow product a t 69-71" (0.5mm.), n y 1.4948.When this material waa chilled a t 0" for several hours, it crystallized to colorless blades which melted at 24'.
Acknawledgment. The authors wish to thank Alfred Foulds for the microanalyses, and C. F. Hartman for technical assistance in this work. PHILLIPSBURQ. N. J.
RESEARCH LABORATORIES, TENNESSEE EASTMAN CO., DIVISIONOF EASTMAN KODAKGO.]
Chemistry of Dimethylketene Dimer. I. Catalytic Hydrogenation and Ring Cleavage by Alcohols ROBERT H. HASEK, EDWARD U. ELAM, JAMES C. MARTIN, AND RONALD G. NATIONS
Received June SO, 1960 The hydrogenation of dimethylketene dimer (I) was investigated to find the optimum conditions for preparation of the corresponding glycol 111. Excellent yields were obtained with a ruthenium catalyst. Hydrogenation over Raney nickel was often accompanied by formation of by-producte. In a methanolic medium, some of the dimer was cleaved to form methyl 2,2,4trimethyl-3-oxovalerate(VI). This reaction, when catalyzed by basea, was generally applicable to other alcohols and to phenols and mercaptans. Another by-product was identified as l-hydroxy-2,2,4-trimethyl-3-pentanone(V), which presumably waa formed by ring opening of the intermediate ketol I1 and subsequent hydrogenation of the acyclic keto aldehyde IV formed by this cleavage. The glycol I11 was separated into isomers melting at 148" and 163O,which were characterized by infrared and NMR spectra as the trans and cis isomers, respectively.
Tetramethyl-l,3-cyclobutanedione (I) was fmt prepared by Wedekind and Weisswange' by the dehydrohalogenation of isobutyryl chloride with triethylamine. At about this same time, Staudinger and Kleverz succeeded in preparing dimethylketene from bromisobutyryl bromide and zinc, and it was recognized that this compound spontaneously formed a dimer which was identical with I.
Zn
(CH&CBrCOBr
t
+(CHs)&=C=O
mann rearrangement of the Wolff-Kishner reduction of the disemicarbazone,aand electron diffraction studies.' An unsuccessful attempt was made to obtain the corresponding glycol, 2,2,4,4tetramethyl-l,3-cyclobutanediol (111)) by reduction of I with sodium amalgam.6 Catalytic hydrogenation under relatively mild conditions produced the intermediate ketol, 3-hydroxy-2,2,4,4-tetramethylcyclobutanone (11). Gross' first carried out this reduction over platinum oxide catalysts in equipment similar to the Adams-Voorhees apparatus: but Miller' obtained better results (80% yield) with Raney nickel under 5-6 atm. pressure of hydrogen, More drastic conditions, up to 150' and 100 atm. over Raney nickel, were used by Miller to force the hydrogenation to the glycol 111.
(4) L.L. Miller, Structure of Some Derivatives of DimethylThe structure of dimethyketene dimer as a cyclic ketene, Ph.D. thesis, Cornell University, 1037. 1,3diketone has been well established by the for(5) W. N. LiDacomb and V, Schomaker. J . C!hem. Phzis.. mation of typical carbonyl derivatives, the Eeck- 14; 475 (1946): (6) E.Wedekind and M. Miller, Ber., 44,3285 (1911). (7) P. F. Gross, The Structure of Ketene Dimer, Ph.D. (1) E. Wedekind and W. Weisswange, Ber., 39, 1631 -
,
lts
(1906). (2) H. Staudingcr and H. W. Klever, Bet., 39, 968 (1906). (3) H. L.Hcrzog and E. R. Buchman, J . Org. Chem., 16, 99 (1951).
thesis, Cornell University, 1936. (8) R. Adams, V. Voorhees, and R. L. Shriner, Org. Syntheses, Coll. Vol. I, 463 (1941). * (9) R. A d a m and V. Voorhees, Org. Syntheses, Coll. V01n I, 61 (1941).
